Communication pubs.acs.org/crystal
Effects of Additives on Crystallization in Thin Organic Films Thomas R. Fielitz, Christopher M. Phenicie,† and Russell J. Holmes* Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota 55455, United States S Supporting Information *
ABSTRACT: Controlling the shape and growth of crystals in molecular organic solids has ramifications impacting diverse fields, but remains challenging to fully exploit. Here, crystal shapes in organic thin films are manipulated from aspect ratios of 1 to over 50, with corresponding growth rates decreased by an order of magnitude simply by mixing a structurally dissimilar minority species into the film. These effects are mapped with composition and temperature in mixtures of two model small-molecular-weight organic compounds, revealing a continuous variation in crystal shape and growth rate. Other combinations of molecules are discussed, showing additive shape selection in multicomponent mixtures and enabling customization of crystal shape.
S
thickness- and temperature-dependent transitions in polymorph composition and growth mode during annealing.28−31 In particular, pure rubrene grows in equiaxed, single-crystalline grains of the orthorhombic polymorph at temperatures 160 °C < T < 180 °C as shown in Figure 1a. These crystals are characterized by twofold symmetry, featuring a faceted vertex defined by the {110} planes along the [010] direction and a rounded edge formed by competition between the (100) and {110} planes intersecting the [100] direction, while at higher temperatures the orthorhombic polymorph grows as a spherulite.28 TPBi is a glass-forming molecule which crystallizes upon annealing and has greater steric bulk than rubrene, making it ideal for studying the disruption of rubrene crystal growth and vice versa. With the inclusion of TPBi, the rubrene crystal shape changes dramatically: as the concentration of TPBi is increased, the crystals show pronounced elongation along the [010] direction, effectively becoming microwires at concentrations above 20 vol % TPBi (Figure 1b−f). All listed concentrations are for the as-deposited films, prior to annealing. The transition in shape between 15 and 20 vol % TPBi is preceded by the change observed in 15 vol % films at high temperature (Figure 1d), where growth proceeds primarily along the [010] direction, resulting in many narrow crystals. Note that the [010] direction is also the direction of highest hole mobility in rubrene,27 convenient for possible use as microwire transistors. A shape change is also observed in TPBirich crystals, where the neat films show equiaxed, hexagonal grains which become elongated along the acutely faceted direction with the addition of rubrene (Figure 1g−i). We note that growth of smooth crystals in TPBi films has not been
olid-state mixtures of organic molecules are found in a wide variety of applications including organic electronic devices and pharmaceuticals.1−3 In addition to sought-after mixture effects such as enhanced excited state dissociation,4,5 charge conduction,6 or drug bioavailability,7,8 mixtures can alter the fundamental behavior of their constituents in unforeseen ways, including their crystallization.9−11 Directed control of crystal size and shape has been well-characterized in solution-grown inorganic nanoparticles12−14 and nanowires15 and has been described in organic crystals grown from solution.16−18 However, this manipulation remains elusive in the solid state, where the ability to direct crystal shape is highly desirable for patterning organic field-effect transistors (OFETs).19,20 Similarly, the effects of macromolecular additives, e.g., polymer excipients, on crystal growth rates in pharmaceutical solids and melts are well-known,11,21−23 if still not fully understood, but studies of facet-dependent growth or mixtures of smallmolecular-weight species are lacking. These conditions, relevant to drug mixtures and organic light-emitting devices (OLEDs), are important to gain additional understanding of methods to halt crystallization for enhanced stability of pharmaceuticals and OLEDs,24−26 each of which often comprise highly conjugated organic molecules in amorphous mixtures. The present work reports tunable crystal shape from discs to needles and conditions to stifle crystal growth in annealed solids via simple mixing, demonstrated here the use of thin films of two archetypal small-molecular-weight organic semiconductors. The molecules used in this study are rubrene (5,6,11,12tetraphenyltetracene), a common active material in OFETs,27 and 2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H-benzimidazole) (TPBi), a common electron transport material in OLEDs.2 The high hole mobility of rubrene has driven exhaustive research into understanding its crystallization from amorphous thin films, where rubrene exhibits complicated © XXXX American Chemical Society
Received: June 22, 2017 Revised: August 16, 2017
A
DOI: 10.1021/acs.cgd.7b00880 Cryst. Growth Des. XXXX, XXX, XXX−XXX
Crystal Growth & Design
Communication
Figure 2. Effects of additive location on crystal shape in 20-nm-thick annealed rubrene films, including linearly graded films with the composition varied from (a) 20−0 vol % TPBi, (b) 0−20 vol % annealed at 180 °C, and uniformly mixed 35 vol % TPBi films at (c) 53, (d) 71, and (e) 75 s and 210 °C. Concentration profiles plotted in (a) and (b) are as-deposited, prior to annealing. White arrows and dotted lines highlight the new growth of dark and light TPBi-rich phases. All images are polarized optical micrographs.
plateauing between 60 and 80 vol % TPBi and correspondingly longer times required for crystals to nucleate and grow (Figure 3a,b), presumably delayed by the loss of rubrene to a critical
Figure 1. Crystal shapes at different compositions (vol % TPBi in asdeposited film) and annealing temperatures: (a) 0, 160 °C; (b) 10, 170 °C; (c) 15, 170 °C; (d) 15, 190 °C; (e) 20, 170 °C; (f) 25, 170 °C; (g) 80, 200 °C; (h) 90, 190 °C; and (i) 100, 190 °C. The inset molecular structures are (a) rubrene and (i) TPBi. Annealing times in seconds are inset in the upper-right corner of each image, and all images are polarized optical micrographs. (a−f) are 20 nm thick, (g−i) are 30 nm thick.
previously observed, making this result noteworthy for crystalline devices. Such behavior has also been found in films of other archetypal organic electronic materials during the course of this work (Figure S1). For films in which each species has sufficiently diluted the other (between 40 and 60 vol % TPBi), neither crystal type is able to efficiently nucleate or grow, causing crystal growth to be stunted. Majority-rubrene films simply do not grow sizable rubrene crystals, while majority-TPBi films require substantial rubrene sublimation to before growth onset. To assess where this shape changing effect originates, the relative placement of the additive within the film was varied by growing films of graded composition shown in Figure 2. Interestingly, the shape produced does not depend strongly on additive location. Regardless of whether the TPBi-rich interface is at the substrate (Figure 2a) or top film surface (Figure 2b), the crystal shapes resemble the case of uniformly mixed films with the average gradient composition. This suggests that bulk diffusion, albeit considerably slower than surface diffusion,32 is significant for these films at the annealing temperature. A complication of this is that films may change in composition over time as the more volatile species (i.e., rubrene) sublimes from the film surface, as in highly concentrated but majorityrubrene films at relatively high temperatures. These films exhibit slow, stagnating rubrene needle growth at short times (Figure 2c), followed by secondary growth of TPBi-rich domains at considerably longer times (Figure 2d,e). This implies that at long times, there is appreciable evaporation of rubrene during annealing at high temperatures, effectively concentrating the films to the point that TPBi crystals may grow. Films with high concentrations of TPBi also show the effects of rubrene sublimation, with the crystal aspect ratio
Figure 3. Properties of rubrene films with varying TPBi concentration in the as-deposited film: (a) maximum crystal aspect ratio, (b) nucleation time, tNuc, and (c) growth rate along the elongated crystal axis. Majority-rubrene films are 20 nm thick; majority-TPBi films are 30 nm thick.
concentration. Majority rubrene films instead show a reduction in aspect ratio at high concentrations since the crystals are unable to grow quickly enough before the surrounding film sublimes. In addition to these observations, thermogravimetric data demonstrate that rubrene is considerably more volatile than TPBi (Figure S2),33 which reinforces the proposal of significant mass loss of primarily rubrene with annealing. If developed further, these observations may provide a unique method to determine the relative rates of diffusion and sublimation for a material within a crystallizing film. B
DOI: 10.1021/acs.cgd.7b00880 Cryst. Growth Des. XXXX, XXX, XXX−XXX
Crystal Growth & Design
Communication
Elongation of crystal shape represents an alteration in crystal planar growth rates from the pure film case, with possible causes including facilitated growth of the long axis,34−36 impeded growth of the short axis,12,13 or some combination thereof. Examination of the crystal growth rate with concentration and temperature (Figure 3c) shows there is a decrease in growth rate of the elongated axis with mixing, but an even more pronounced reduction in growth rate along the short axis, as can be inferred from the increase in crystal aspect ratio. Figure 3c implies that the crystal shape is governed by secondary molecules impeding the attachment of primary, crystal-forming molecules rather than enhancing their attachment, very similar to what is observed in the shape-controlled synthesis of solution-grown nanoparticles.12,13 It is unclear whether this change of rate corresponds to growth delay from additive exclusion, lattice disruption by inclusion of additive pockets, or other mechanisms previously discussed.22,23 If the growing crystal excludes material, as with rigorously stoichiometric inorganic mixtures,36−38 the minority species could be expected to partition to the crystal surfaces or become included in the bulk as a concentrated additive cluster. Despite the multiple molecular conformations and weak van der Waals bonds common in polymorphic organic crystals, perhaps increasing susceptibility to additive incorporation, all collected X-ray diffraction (XRD) patterns on mixed films do not deviate from pure-material spectra (Figure S3). This suggests that the crystalline lattice is unaltered by the additive, but does not preclude either of the proposed mechanisms. As the TPBi content increases in majority-rubrene films, the rubrene attachment rate is significantly slowed along both crystal directions, but especially along the [100] direction. Eventually, the rate of [100] growth becomes unable to compete with that of sublimation from the film, causing material near the crystal edge to move preferentially via a vapor cloud over the film and deposit on the top surface of the crystal edge rather than continue lateral growth. This leads to a trough-like appearance of the needles with atomic force microscopy (AFM, Figure 4a), as well as the formation of a material depletion region along the resulting (100) sides of the crystals, essentially ceasing growth and preventing other crystals from growing in the vicinity (Figure 4b). Due to the higher rate of [010] growth, the [010] direction stays competitive with material diffusion, allowing growth to continue. Examination of quenched films shows no visible gap at the needle tip ([010]), with marked depletion of material along the needle sides ([100]) frustrating lateral crystal growth (Figure S4). This growth asymmetry causes the crystals to continuously elongate until the ultimate stagnation of the [010] direction, likely after TPBi concentration from rubrene sublimation slows lateral growth below that of vapor transport. Neat rubrene is known to undergo a transition from singlecrystal to polycrystalline spherulitic growth of the orthorhombic polymorph with temperature,28 which is also observed in rubrene-rich mixed films. At TPBi concentrations below 15− 20% (below the needle-like growth range), the high-temperature mixed rubrene grains appear very similar to neat rubrene. In the case of the needle-like crystals, the transition to spherulitic growth is accompanied by macroscopic, optically resolvable branching (Figure 4c,d). As the annealing temperature increases, the branching frequency increases as well, resulting in a finely grained spherulite at low TPBi concentrations and an interwoven web of crystals at high concentrations.
Figure 4. Topography and temperature dependence of 20-nm-thick 20 vol % (as-deposited) TPBi films. (a) Atomic force micrograph of a crystal grown at 190 °C (logarithmic height scale, 296 nm), (b) unpolarized optical micrograph of light-colored depleted regions around growing crystals in a film quenched after annealing at 180 °C, and polarized optical micrographs showing branching of crystals grown at (c) 185 °C and (d) 200 °C.
The effects described thus far show the interaction of rubrene and TPBi, but other molecules also have an effect on the crystal growth of rubrene, as shown in Figure 5. From the selection of molecules presented, it is apparent that the axis of influence for
Figure 5. Polarized optical micrographs of crystal shapes of 20-nmthick rubrene films doped with (a) 30 vol % α-NPD at 190 °C, (b) 15 vol % 4p-NPD at 180 °C, (c) 10 vol % TCTA at 180 °C, and (d) 10 vol % TCTA and 10 vol % TPBi at 180 °C. All concentrations are prior to annealing. C
DOI: 10.1021/acs.cgd.7b00880 Cryst. Growth Des. XXXX, XXX, XXX−XXX
Crystal Growth & Design
Communication
TPBi is not universal; however, inducing shape change in rubrene crystals is. That similar molecules like α-NPD (N,N′di(1-naphthyl)-N,N′-diphenyl-(1,1′-biphenyl)-4,4′-diamine, Figure 5a) and 4p-NPD (N,N′-di-1-naphthalenyl-N,N′-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴-diamine, Figure 5b) produce similar shape modifications in rubrene suggests perhaps shape manipulation is related to the molecular functional units (e.g., binding or miscibility) or rigidity rather than strictly molecule size. Either parameter could influence how an additive molecule hinders lattice site attachment in a growing crystal or diffuses to aggregate elsewhere. If the determining factor were strictly size, a bulky molecule like TCTA (tris(4-carbazoyl-9-ylphenyl)amine) might be expected to behave more similarly to TPBi than α-NPD, which is not obvious from the collected data (Figure 5c). A final dimension to these alternative additives is the incorporation of multiple species, such as TCTA and TPBi, as shown in Figure 5d. In this example, the crystal shape appears to resemble an elongated cross-shaped structure; a hybrid between the shapes observed in Figure 1b and Figure 5c. This adds another axis by which to customize crystal shape selection in thin organic films. Further work more closely examining crystal growth with a series of impurities systematically varied in stiffness, size, and intermolecular interaction would be helpful in elucidating the underlying mechanism. This work opens or furthers several avenues of research. First, the ability to create precisely defined regions of crystalline material with depleted regions on either side could aid in developing OFETs with low degrees of conduction between separate transistors, crucial for scaling down devices. Second, monitoring crystal shape and growth rate allows for optical measurement of molecular concentrations in thin films. This presents a new method to study solid-state diffusion of molecular species through thick or buried layers, using materials which are relevant for electronic devices with device-relevant geometries and spatial scales. This approach could also be applied to study pharmaceutical compounds when deposited in thin-film geometries.39,40 Finally, to build on work with polymer additives in pharmaceuticals,11,21,22 using mixed films of small molecules to understand and control crystallization can be further quantified and exploited to improve the stability of drugs and electronic devices. With these possibilities come challenges, such as controlling crystal nucleus location and orientation in OFETs, ensuring crystal growth of a layer deposited atop an arbitrary material stack, and maintaining optoelectronic and medical properties of a material while incorporating increasing amounts of a secondary species; however, these opportunities are enticing. In conclusion, mixed thin films of small organic molecules, specifically rubrene and TPBi, produce dramatically varying crystal shapes and growth rates when annealed. This variation results from the impediment of crystal growth by the film minority species to differing degrees for different crystallographic axes, with additional effects due to film sublimation yielding surprisingly elongated crystals for films with a high loading of additive. The resulting crystal shape varies depending upon the additive type, with cumulative shape-changing properties for ternary films.
■
■
Experimental procedures, thermogravimetric data, X-ray diffraction patterns, atomic force micrographs (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Russell J. Holmes: 0000-0001-7183-3673 Present Address
† Department of Electrical Engineering, Princeton University, Princeton, NJ 08544.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Science Foundation (NSF) Solid State and Materials Chemistry (SSMC) Program under DMR-1307066.
■
REFERENCES
(1) Byrn, S.; Pfeiffer, R.; Ganey, M.; Hoiberg, C.; Poochikian, G. Pharm. Res. 1995, 12 (7), 945. (2) Tao, Y.; Yang, C.; Qin, J. Chem. Soc. Rev. 2011, 40, 2943. (3) Peumans, P.; Uchida, S.; Forrest, S. R. Nature 2003, 425, 158. (4) Hiramoto, M.; Fujiwara, H.; Yokoyama, M. Appl. Phys. Lett. 1991, 58 (10), 1062. (5) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Science 1995, 270 (5243), 1789. (6) Schubert, S.; Kim, Y. H.; Menke, T.; Fischer, A.; Timmreck, R.; Müller-Meskamp, L.; Leo, K. Sol. Energy Mater. Sol. Cells 2013, 118, 165. (7) Vasconcelos, T.; Sarmento, B.; Costa, P. Drug Discovery Today 2007, 12, 1068. (8) Serajuddin, A. T. M.; Sheen, P.-C.; Augustine, M. A. J. Pharm. Sci. 1990, 79 (5), 463. (9) Zhang, J.; Li, C.-Z.; Williams, S. T.; Liu, S.; Zhao, T.; Jen, A. K.-Y. J. Am. Chem. Soc. 2015, 137, 2167. (10) Winn, D.; Doherty, M. F. AIChE J. 2000, 46 (7), 1348. (11) Sun, Y.; Zhu, L.; Wu, T.; Cai, T.; Gunn, E. M.; Yu, L. AAPS J. 2012, 14 (3), 380. (12) Guo, S.; Fidler, A. F.; He, K.; Su, D.; Chen, G.; Lin, Q.; Pietryga, J. M.; Klimov, V. I. J. Am. Chem. Soc. 2015, 137, 15074. (13) Xia, Y.; Xia, X.; Peng, H.-C. J. Am. Chem. Soc. 2015, 137, 7947. (14) Polavarapu, L.; Mourdikoudis, S.; Pastoriza-Santos, I.; Pérezjuste, J. CrystEngComm 2015, 17, 3727. (15) Salvalaglio, M.; Backofen, R.; Bergamaschini, R.; Montalenti, F.; Voigt, A. Cryst. Growth Des. 2015, 15, 2787. (16) Lovette, M. A.; Browning, A. R.; Griffin, D. W.; Sizemore, J. P.; Snyder, R. C.; Doherty, M. F. Ind. Eng. Chem. Res. 2008, 47 (24), 9812. (17) Borsos, A.; Majumder, A.; Nagy, Z. K. Cryst. Growth Des. 2016, 16, 555. (18) Huang, L.; Liao, Q.; Shi, Q.; Fu, H.; Ma, J.; Yao, J. J. Mater. Chem. 2010, 20, 159. (19) Briseno, A. L.; Roberts, M.; Ling, M.-M.; Moon, H.; Nemanick, E. J.; Bao, Z. J. Am. Chem. Soc. 2006, 128, 3880. (20) Liu, S.; Wang, W. M.; Briseno, A. L.; Mannsfeld, S. C. B.; Bao, Z. Adv. Mater. 2009, 21, 1217. (21) Cai, T.; Zhu, L.; Yu, L. Pharm. Res. 2011, 28, 2458. (22) Kestur, U. S.; Lee, H.; Santiago, D.; Rinaldi, C.; Won, Y.; Taylor, L. S. Cryst. Growth Des. 2010, 10 (8), 3585. (23) Aso, Y.; Yoshioka, S.; Kojima, S. J. Pharm. Sci. 2004, 93 (2), 384. (24) Tyagi, P.; Srivastava, R.; Giri, L. I.; Tuli, S.; Lee, C. Synth. Met. 2016, 216, 40. (25) Fenter, P.; Schreiber, F.; Bulovic, V.; Forrest, S. R. Chem. Phys. Lett. 1997, 277, 521. (26) Lee, E. H. Asian J. Pharm. Sci. 2014, 9 (4), 163.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.cgd.7b00880. D
DOI: 10.1021/acs.cgd.7b00880 Cryst. Growth Des. XXXX, XXX, XXX−XXX
Crystal Growth & Design
Communication
(27) Sundar, V. C.; Zaumseil, J.; Podzorov, V.; Menard, E.; Willett, R. L.; Someya, T.; Gershenson, M. E.; Rogers, J. A. Science 2004, 303, 1644. (28) Fielitz, T. R.; Holmes, R. J. Cryst. Growth Des. 2016, 16 (8), 4720. (29) Lee, H. M.; Moon, H.; Kim, H.-S.; Kim, Y. N.; Choi, S.-M.; Yoo, S.; Cho, S. O. Org. Electron. 2011, 12, 1446. (30) Käfer, D.; Witte, G. Phys. Chem. Chem. Phys. 2005, 7, 2850. (31) Park, S.-W.; Choi, J.-M.; Lee, K. H.; Yeom, H. W.; Im, S.; Lee, Y. K. J. Phys. Chem. B 2010, 114 (17), 5661. (32) Zhu, L.; Brian, C. W.; Swallen, S. F.; Straus, P. T.; Ediger, M. D.; Yu, L. Phys. Rev. Lett. 2011, 106, 256103. (33) Price, D. M. Thermochim. Acta 2001, 367−368, 253. (34) Hempel, T.; Schoenfeld, O.; Syrowatka, F. Solid State Commun. 1993, 85 (11), 921. (35) Nygren, E.; McCallum, J. C.; Thornton, R.; Williams, J. S.; Olson, G. L. MRS Online Proc. Libr. 1988, 100, 405. (36) Nygren, E.; Williams, J. S.; Pogany, A.; Elliman, R. G.; Olson, G. L.; McCallum, J. C. MRS Online Proc. Libr. 1986, 74, 307. (37) Jacobson, D. C.; Elliman, R. G.; Gibson, J. M.; Olson, G. L.; Poate, J. M.; Williams, J. S. MRS Online Proc. Libr. 1986, 74, 327. (38) Radnóczi, G. Z.; Knez, D.; Hofer, F.; Frangis, N.; Vouroutzis, N.; Stoemenos, J.; Pecz, B. J. Appl. Phys. 2017, 121, 145301. (39) Rodríguez-Tinoco, C.; Gonzalez-Silveira, M.; Ràfols-Ribé, J.; Lopeandia, A.; Rodriguez-Viejo, J. Phys. Chem. Chem. Phys. 2015, 17, 31195. (40) Swallen, S. F.; Kearns, K. L.; Mapes, M. K.; Kim, Y. S.; McMahon, R. J.; Ediger, M. D.; Wu, T.; Yu, L.; Satija, S. Science 2007, 315, 353.
E
DOI: 10.1021/acs.cgd.7b00880 Cryst. Growth Des. XXXX, XXX, XXX−XXX
